A Method for Parametric and Catastrophic Fault Diagnosis of Analog Linear Circuits

Abstract

A universal method for single parametric and catastrophic fault diagnosis of analog linear circuits is presented in this paper. The methodology is based on fundamental laws governing linear circuits and methods of their analysis. The method involves creating models of the faulty elements, both passive and active, including current and voltage sources and applying an appropriately modified node method. This enables the creation of simple formulas to define the parameters of the faulty elements. Some measurement data must be collected during the course of the diagnostic test performed in the frequency domain and certain computation results obtained in the before-test-process. The method achieves all of the objectives of fault diagnosis: detection, location, and estimation of the faulty value. In 88.48%, the method correctly identifies the fault and estimates its value, but in 6.47%, the actual fault is accompanied by a virtual fault. The method is adapted to real conditions to improve the practical relevance and can be directly extended to double fault diagnosis. Six numerical examples are presented to illustrate the method.

Analog circuit diagnosis based on support vector machine with parameter optimization by improved NKCGWO

“…widely used and play a vital role in various electronic systems. An unexpected failure of analog circuit may lead to sudden breakdown of the entire equipment, resulting in huge economic losses or even casualties [1,2,3]. However, analog fault diagnosis is still a challenging task because of poor fault models, component tolerances and nonlinear effects of analog circuits [4,5].…”